Exploring the Edge — Observations from our Hackathon with SKT

August 21st, 2019

Headshot for Wonho Park

Wonho Park

Chief Product Officer & SVP Product Development

As previously reported, earlier this month the MobiledgeX team helped to host a 5G and edge computing hackathon with SKT, the largest mobile operator in South Korea, to give developers hands-on access to 5G and the mobile edge. Competitive challenges allowed them to explore and better understand the benefits of these next generation technologies. 

I had the opportunity to travel to Bundang-gu, South Korea, where I attended the hackathon and interacted directly with each team of developers. With the challenges running on our MobiledgeX Edge-Cloud console, I was eager to see how developers would harness the power of 5G and edge computing based on their respective areas of expertise, ranging from industrial IoT to AI/ML. Pre-hackathon surveys revealed the developers were mostly very early in their consideration of edge, so I was prepared to help educate and inspire. 

Packed house at the SKT Hackathon

It was a packed house and the energy was palpable, a combination of comradery and competitive spirit. 

As I moved around the room, from team to team, common questions and objectives quickly became clear. Developers were looking for the “ability to leverage geo-location data to create applications that adapt based on user location,” were excited about “reducing the amount of data traffic and latency that has to go through the network by processing it closer to the device,” and were “interested in running high compute workloads to an edge site” (often described as compute offload). For those of you who know the combined capabilities of 5G and edge computing, you’re probably thinking these developers showed up to the right place. I agree. 

The biggest challenge developers faced was not understanding the technology, but rather understanding and exploring how the technology could offer new end user experiences. 

KPST, the winning team, did just that: they reimagined what’s possible using edge-augmented image processing. They created an image detection application that calculates calories (called Diet App). Their current version of the app runs a detection algorithm (a Machine Learning algorithm) on the local device. Limited processing power on the device impacted the performance of the algorithm available. By leveraging 5G mobile edge computing, the team from KPST was finally able to reimagine their app and improve accuracy and user features and functionality. 

KPST, the winning team

The biggest lesson we (MobiledgeX, SKT, and the other sponsors including Intel, Samsung, and Microsoft) learned is that developers have many use cases today that could benefit from 5G and edge computing, but there is little awareness on the part of developers as to how to do this in reality. Similarly, there is little awareness on the part of technology providers as to the breadth and specifics of developer use cases. Events like this hackathon help to move the conversation around 5G and edge computing forward so that all parts of the ecosystem can benefit.  

With so much high-level and abstract hype surrounding 5G and edge computing, it was refreshing to see developers roll up their sleeves and actually build on these technologies.

Many thanks to SKT for hosting this groundbreaking event. It was an honor to be involved.